Vendax LipSync Studio

Vendax LipSync Studio — User guide

Purpose

Vendax LipSync Studio adapts lip movements in a video or portrait photograph to a prepared voiceover. Use it for dubbing, language versions and talking portraits. It does not translate text or generate voices. Processing runs locally once the required components and models are installed.

Workflow

  1. Save the project. On Edit, place a video or photograph on a video track and the new speech on a separate audio track, such as A1. Align the speech with the intended section. Set the photograph’s duration. Keep music on another track.
  2. Open Workspace → Workflow Integrations → Vendax LipSync Studio. The screenshot shows Deliver: switch to Edit and select exactly one clip on a video track.
  3. Click Refresh. Under Selected Video Clip, check the clip name and project / timeline.
  4. In New Voiceover Track, select the existing speech track. This menu does not create a track. Audio must overlap the processed section.
  5. In AI Engine, choose an installed processing engine. Start with a five-second preview. A single clearly visible person with an unobstructed mouth is recommended. Review the result before processing the entire clip.
  6. First leave Preview: first 5 seconds of the clip enabled. It processes the first five seconds, or the entire clip if shorter. This is sample length, not processing time.
  7. Click Sync Lip Movements. During preparation, Resolve temporarily opens a duplicate timeline and Deliver. Wait for the original timeline to return. Do not switch projects or modify the source during processing. Use Cancel to stop and wait for cancellation.
  8. When Ready appears, click Add to Timeline. Output appears at the source clip’s start on a new top video track, Vendax LipSync Studio. The original remains and the voiceover stays on its audio track. The output clip contains video only.
  9. Review lips, face edges, transitions and speech timing. Disable the output track temporarily to compare. For a full render, uncheck Preview, reselect the original, click Refresh and run again. Disable or remove the test layer from the timeline so it cannot cover the final output. Save the project.

Troubleshooting

Sync Lip Movements disabled: check Edit, select one video clip, ensure an audio track exists and click Refresh. No audio clips overlap…: check track selection and alignment; the preview needs audio in the first five seconds. Add to Timeline disabled: wait for readiness and open the original project and timeline. If the source moved or changed, process again. Wait for any existing Resolve render. If required components or models are missing, repair the panel installation. If memory is insufficient, close other applications and start with a five-second test.

Keep your output

Results are stored under engine/MuseTalk/jobs/job_…/result.mp4 in the project folder. Do not move or delete job folders while Resolve references them: media can go offline. Saving the Resolve project does not replace keeping its media. Include output media when transferring the project.